All depends on costs. We priced out 2x10G vs 1x100G between two datacenters, and with the XC charges on both sides of the pipe for 2x10G, the 1x100G came in cheaper.
If you did “B”, you could light your own 100G for the cost of a 9 mile run + dark fiber lease. > On Aug 26, 2020, at 12:32 PM, Mark - Myakka Technologies <[email protected]> > wrote: > > We are starting to get close enough on upstream bandwidth, where I > need to start thinking about getting some more. > > Currently we have two 10G ports going to two different data centers. > With the internet connections, IX links and current hardware, I can > pull about 8GB max from each port. Currently in peak times we pull a > total of about 4-5 gbps on one line and about 3-4 gbps on the other. > It is manageable now, until an upstream goes down during peak times. > That has only happened once and it was a bit dicey. > > The question is which would be "better"? > > A. Upgrade the two pipes to 20/40/100G whatever is the standard now. > Buy new hardware to handle the new speeds needed. > > B. Build out a 9 mile fiber run to a possible POP where I can > possibly grab some dark fiber or at least another 10G and run to a possible > 3rd data center. > > I personally like B, being I can push upgrading the other two links > down the road. This is a rural area, but a state road, so the fiber > install won't be cheap. However, I can possibly pick up 10-15 > customers along the way with the possibility of getting more as things > build out. > > Currently the two upstreams balance themselves out. I don't > have any fancy code doing load balancing. If I add a third to the > mix, I'm not sure how well they will balance out. > > Just trying to figure out pros and cons of each. > > > -- > > Thanks, > Mark mailto:[email protected] > > Myakka Technologies, Inc. > www.Myakka.com > > > -- > AF mailing list > [email protected] > http://af.afmug.com/mailman/listinfo/af_af.afmug.com -- AF mailing list [email protected] http://af.afmug.com/mailman/listinfo/af_af.afmug.com
